Closed-Loop Low-Rank Echocardiographic Artifact Removal.

Echocardiographic image sequences are frequently corrupted by quasi-static artifacts ("clutter") superimposed on the moving myocardium. Conventionally, localized blind source separation methods exploiting local correlation in the clutter have proven effective in the suppression of these artifacts. These methods use the spectral characteristics to distinguish the clutter from tissue and background noise and are applied exhaustively over the data set. The exhaustive application results in high computational complexity and a loss of useful tissue signal. In this article, we develop a closed-loop algorithm in which the clutter is first detected using an adaptively determined weighting function and then removed using low-rank estimation methods. We show that our method is adaptable to different low-rank estimators, by presenting two such estimators: sparse coding in the principal component domain and nuclear norm minimization. We compare the performance of our proposed method (CLEAR) with two methods: singular value filtering (SVF) and morphological component analysis (MCA). The performance was quantified in silico by measuring the error with respect to a known "ground truth" data set with no clutter for different combinations of moving clutter and tissue. Our method retains more tissue with a lower error of 3.88 ± 0.093 dB (sparse coding) and 3.47 ± 0.78 (nuclear norm) compared with the benchmark methods 8.5 ± 0.7 dB (SVF) and 9.3 ± 0.5 dB (MCA) particularly in instances where the rate of tissue motion and artifact motion is small (≤0.25 periods of center frequency per frame) while producing comparable clutter reduction performance. CLEAR was also validated in vivo by quantifying the tracking error over the cardiac cycle on five mouse heart data sets with synthetic clutter. CLEAR reduced the error by approximately 50%, compared with 25% for the SVF.

Dictionary learning-based reverberation removal enables depth-resolved photoacoustic microscopy of cortical microvasculature in the mouse brain.

Photoacoustic microscopy (PAM) capitalizes on the optical absorption of blood hemoglobin to enable label-free high-contrast imaging of the cerebral microvasculature in vivo. Although time-resolved ultrasonic detection equips PAM with depth-sectioning capability, most of the data at depths are often obscured by acoustic reverberant artifacts from superficial cortical layers and thus unusable. In this paper, we present a first-of-a-kind dictionary learning algorithm to remove the reverberant signal while preserving underlying microvascular anatomy. This algorithm was validated in vitro, using dyed beads embedded in an optically transparent polydimethylsiloxane phantom. Subsequently, we demonstrated in the live mouse brain that the algorithm can suppress reverberant artifacts by 21.0 ± 5.4 dB, enabling depth-resolved PAM up to 500 µm from the brain surface.